"Mental health services are of no value to patients and the larger community unless they are used by the persons who need them". This is of particular concern where adolescents with multiple alcohol, drug and mental (ADM) disorders are concerned, because (1) they are the responsibility of many agencies; and (2) they rarely refer themselves for help, and may go unrecognized for a long time. The proposed study will examine service use by adolescents with ADM disorders, in a rural area of the southern Appalachians. The study will be carried out as Public Academic Liaison with the Division of Mental Health, Developmental Disabilities and Substance Abuse Services of the North Carolina Department of Human Resources. Adapting Andersen's model of medical care utilization to services for adolescent ADM problems, we propose to describe service use in terms of (1) the community, (2) the family, and (3) the individual. Community factors include rural/urban settings and the distribution of services there. Family factors may be a source of risk or protection: they include family psychiatric history and functioning, income, insurance, knowledge and beliefs about adolescent ADM services, and demographic and social characteristics. Individual factors include ADM disorders and comorbidity, functioning in major roles at school, home, and work, and attitudes to service use. Our concern is how these factors interact over time to predict service use, as well as range of real-world outcomes such as school dropout, arrests, accidents, and pregnancies. We shall examine how the amount of service use (type of care, restrictiveness, intensity, duration) is influenced over time by the interaction of ADM comorbidity with other individual, family and community factors. The findings will point to areas where developments in treatment, family intervention, and service planning can have maximum effectiveness. The proposed study is set in 11 counties in western N. Carolina, bordering Tennessee. The area is predominantly rural, but contains one town of 60,000. A probability sample will be drawn, from data provided by the Department of Public Instruction, made up to 3 cohorts aged 10, 12.5, and 15. Following telephone screening to oversample those at risk for behavioral or substance abuse problems, 50% (1,500) will recruited into the main sample. Parent/caretaker and child will be interviewed three times, at yearly intervals. Data will be obtained from teachers, and from a wide range of service settings. Between interviews, use of services and new ADM symptoms will be reported in a brief telephone interview every three months. Measures include a newly developed instrument for self- and parent-report of services used, and attitudes toward them (the Child and Adolescent Services Assessment: CASA). The CASA covers both formal and informal service systems, and also provides information on perceived benefits, methods of payment, and focus of treatment. Of particular interest is the use made of informal providers of help (e.g., friends, ministers) in rural areas, for problems that may routinely be referred to specialist agencies in urban areas. The proposed study is sited in an area selected for a Robert Wood Johnson demonstration focused on comprehensive services for severely mentally ill youth. A CASSP-funded evaluation of the RWJ project is also ongoing, using the same instruments as the proposed study. These two projects have resulted in the development of (a) a unique identifier for children receiving services, so that multiple service use can be tracked in a manner consistent with confidentiality; and (b) a spirit of cooperation among agencies and with research which makes possible a community study of this kind.